Bach’s use of the single-note ornament in the Goldberg Variations
Paul Badura-Skoda’s article ‘Let’s get rid of the wrong Pralltriller!’1 in the February 2013 issue of
Early Music highlights the fact that after half a cen-tury of performance practice studies, even common elements of ornamentation still demand reassess-ment. This article, written mostly during 2012, con-cerns an even more basic kind of decoration—one which, like the pralltriller, has a performance tradi-tion (particularly in Bach’s music) deriving largely from a few influential performers and scholars. This tradition is hard for performers to ignore. What has come to ‘sound right’ through countless repetitions can be hard to discard, and its replacement by some-thing which might be argued more truly to convey a composer’s intentions, may, at least for a time, be found by many to sound ‘wrong’.
Ornaments consisting of just a single note are the focus of this study. Almost any note in a piece of music can have this most elementary of orna-ments attached to it. As we shall see, it may perform one of several functions: to add stress (either to the following note, or to the ornament itself); to make disjunct motion smoother; and so on. The notation of such simple ornaments was similarly varied: this could range from a truly representative note, written full-size, down to a smaller note (sometimes much smaller), whose performing length was often not defined by the notation alone. Slurs were sometimes added, sometimes not. Even today such ambiguities are not fully understood.
For many keyboard players of J. S. Bach’s time the single-note ornament may have been among the most frequently used of ornaments. But today’s players, using the best modern editions, do not gain this impression when exploring the music of Bach
and his German contemporaries. Compared with a French composer of the period (d’Anglebert and François Couperin are obvious examples) indi-cations seem to be very few, and, of course, this comparison can be extended to ornamentation in general.2 Bach, although more prescriptive than many, and offering what have been termed ‘perfor-mance versions’ of a number of his pieces,3 never-theless often relied, as did others, on a performer’s musicality for the addition of ornaments where none were specified, as well as the appropriate use of those which were indicated.
Bach also had to trust those who copied his music.4 The composer’s circle (his immediate family, friends, admirers and students) played a vital role in preserving and transmitting his music through copying, which often involved alterations—another reason why his intentions may be unclear. In the context of the Goldberg Variations, bwv988, we can, for example, compare the version of the Aria in the first edition with that penned by Anna Magdalena Bach for her own use.5 Anna Magdalena’s presents the notation of single-note ornaments in a way which, were it our only source, would undermine the argument regarding notational differentiation considered below. However, whatever the actual order of events here, it is fair to assume that Bach’s annotation of his own copy of the first edition reflects not only his desire to correct and amplify, but also (where he failed to do so) largely to approve.
The Goldberg Variations contain a number of specified single-note ornaments, mostly but not entirely confined to three movements with a dom-inant melodic line (Aria, Variations 13 and 25). These are worthy of examination for their didactic
pAGE 2 OF 16 Early Music
2.5
2.10
2.15
2.20
2.25
2.30
2.35
2.40
2.45
2.48
2.50
2.55
2.60
2.65
2.70
2.75
2.80
2.85
2.90
2.952.96
potential: they may reflect a performance practice closer to that expected in other music where such indications are few, or non-existent. I feel that an increased use of the single-note ornament could benefit much present-day performance of Bach’s music, enhancing a feeling of spontaneity, elegance of line and emotional force. But there will be no space here for specific recommendations.
This article will discuss the performance style for such ornaments within the Goldberg Variations. The work represents Bach’s most mature thought as music publisher, and is a late work per se. (Even so, his ‘red pen’ corrections and additions to his personal copy show that the edition we have does not offer his final thoughts.) It contains a variety of musical styles, but is predominantly ‘Bachian’—that is, non-derivative and, incidentally, generally non-French—in its inspi-ration. It is also probably his most often-heard large-scale keyboard work today. I will briefly survey the varieties and uses of the single-note ornament, con-centrating upon the Variations themselves. A more extended discussion of the Aria will be given last, since it is in this, both the first and final movement, where radically different treatments are both possi-ble and musically most determinative.
Modern traditionsWe can set aside the choice of some perform-ers to omit many written ornaments altogether when playing Bach.6 Players of the first half of the
20th century who wished to follow a more histori-cally informed route when exploring 200-year-old music, and at least play those ornaments that were provided, naturally sought assistance in the work of editors. One of these, Ralph Kirkpatrick, produced his recommendations at a time that can be consid-ered opportune, since the rediscovery (and, one may add, the reinvention) of the harpsichord was gain-ing momentum. Kirkpatrick’s prescription, given by means of realizations above the score in his edition of 1938,7 helped to establish a 20th-century tradition that has remained dominant to the present day.8 In specific areas (including, importantly, the ornamen-tation of the Aria) very different conclusions had been reached by Edward Dannreuther9 and Arnold Dolmetsch10 decades before the Kirkpatrick edition appeared, and these were to be echoed by Walter Emery in his 1953 book Bach’s ornaments.11 Music, however, is made to be heard. Popular recordings from the 1950s onwards (in particular, increasingly monumental ones by Glenn Gould, but also those by leading figures of the new ‘historical’ school of harp-sichord playing, like Gustav Leonhardt) agreed with key elements of Kirkpatrick’s message, and estab-lished them in the ears of the listening public, and of the majority of pianists and harpsichordists—then and since—as normal and ‘correct’.
Ralph Kirkpatrick’s realization of the ornamen-tation (illus.1) is clear and coherent, with single-note ornaments based generally on proportionate
1 Bach, Goldberg Variations, Aria, bars 1–4, in Ralph Kirkpatrick’s 1938 realization
Early Music PAGE 3 OF 16
3.5
3.10
3.15
3.20
3.25
3.30
3.35
3.40
3.45
3.48
3.50
3.55
3.60
3.65
3.70
3.75
3.80
3.85
3.90
3.953.96
division. The principal source of his interpretation of single-note ornaments is C. P. E. Bach’s Versuch (see Kirkpatrick’s edition, p.xii). It appears that little or no consideration was given to variety of notation. One effect of such a realization is the unfortunate, if natu-ral, tendency of the modern musician to perform it literally (perhaps some of Emanuel’s readers did the same!). This, in turn, will contribute to an enjoyment of the music of Bach for its tone of mathematical purity, rather than for a rediscovery of 18th-century style. It is possible to view the 20th-century approach to much Baroque music (and Bach’s keyboard music in particular), performed in a deliberately unroman-tic manner, as offering an enjoyable antithesis of the ‘heart-on-sleeve’ nature of much Romantic music. But since depth of feeling has always been an essen-tial component of great musicianship, romanticism was here replaced by spirituality: Kirkpatrick’s rec-ommended tempo for the Aria became a little slower with the passage of time.12 Glenn Gould’s recordings of 1955 and 1981 also demonstrate an increasingly introspective search for the spiritual heart of Bach’s genius. Gould increased his extremes of tempo: for example, in his later recording the Aria is played at almost half the speed of the former, causing the mathematical approach to ornamentation to be heard even more clearly, since many ornaments are slowed down proportionately too.13
In the 18th century, ornamentation, whatever the mood of the piece, was essentially the spontane-ous decoration of a line—or it ought in general to sound so. Even Bach, more prescriptive than most, presumably had this in mind when performing and teaching. The 20th-century tradition, at least as exemplified (even if unintentionally) by these two leading and influential exponents, Kirkpatrick and Gould, opposes this.14
The single-note ornament: its forms and notationApart from one hook sign (in bar 15 of the Ouverture, Variation 16), all single-note ornaments indicated by Bach in the first edition of the Goldberg Variations are either written in real note values or indicated by smaller-than-normal notes. With one possible exception (the use of small quavers to indicate only appoggiaturas, which will be considered below) the apparent value of these small notes must not be taken
as a guide to their execution. Most follow an elemen-tary notational ‘grammar’, and are written as notes of half the value of their following notes, but exceptions may easily be found. For example, in bar 2 of the Aria, a small semiquaver precedes a minim; in Variation 13 bar 4, a demisemiquaver precedes a note of the same value; and in one motif repeated throughout Variation 25, small semiquavers precede demisemiquavers.
We cannot know whether Bach wrote any of these as hook signs in his lost manuscript, nor whether he made more than one draft, refining the notation of these ornaments,15 but we will see that the first edi-tion did not constitute his final thoughts. As for the various forms and purposes of the ornament itself, a brief summary must suffice. Although superfluous for many readers, it contains assumptions material to the argument.
The best-known single-note ornament is the appoggiatura. Germans used the neutral terms vorschlag or accent. The French used a variety of terms, and often avoided the inherent false messages that the use of small notes might carry, by the use of more neutral ornament signs. Like some other Germans, Bach followed this practice at times, but his personal ‘hook’ symbol hardly ever appears in his published works, and we can assume that this sign, too, did not dictate any absolute length or placement. The appoggiatura is a stressed ornament, played on the beat, and taking time from the note that follows it. It delays a cadence, and creates dis-sonance against the bass. It can be of varied length, depending on both the context and the player’s own taste, but even when played short, it attracts more emphasis to itself than to the following note. When not written in real note values or by a sign, it is indi-cated by a note of smaller size and value than the note following it, to which it is commonly slurred.
Sometimes the same indication is used for a dif-ferent single-note ornament. This ornament is not stressed, and is played to varying degrees before the beat, taking time from the note that precedes it, and emphasizing that which follows. This orna-ment is called, variously, the grace-note, or passing appoggiatura, or in German nachschlag (literally ‘afterbeat’). Some French authorities positively rec-ommended it, but continued to use the same termi-nology as when discussing the appoggiatura. A full discussion of Bach’s taste (for example, how French
pAGE 4 OF 16 Early Music
4.5
4.10
4.15
4.20
4.25
4.30
4.35
4.40
4.45
4.48
4.50
4.55
4.60
4.65
4.70
4.75
4.80
4.85
4.90
4.954.96
he may sometimes have liked to be) is beyond the scope of this article, but some suggestions will be offered below.
In Germany, towards the end of J. S. Bach’s life, C. P. E. Bach (his second son) and J. J. Quantz (Emanuel’s long-time colleague at the court of Potsdam) both worked in the new galant style, and were in agreement on the musical use both of quite long appoggiaturas, which were often of a specified proportionate nature, and also of very short appog-giaturas, played as far as possible on the beat. In sequences, however, where the purpose of the orna-ment was to link notes (as in intervals of a falling 3rd), their preferences diverged. Quantz enjoyed the orna-ment pulled forward before the beat and de-empha-sized. He called this the ‘Passing Appoggiatura’, and praised it as in the best, French, taste. C. P. E. Bach condemned this practice, while admitting its popular-ity. He preferred sequences of intervals to be linked by real appoggiaturas, which attracted the stress away from the notes that were principal on the page. Both these important figures gave explicit and contrasted realizations of motifs that could easily be given more than one treatment in performance due to the ambig-uous nature of the notation.16
Examples of the types of ornament mentioned above can be found within the Goldberg Variations. Many are unambiguous. Some demand a choice from the performer, made on the basis of deciding the musi-cal affekt that Bach may have had in mind in each case.
The search for J. S. Bach’s tasteAlthough C. P. E. Bach’s treatise The True Art of Playing Keyboard Instruments (the Versuch) of 1753 is regarded as a very important source of insight into his father’s thought (particularly his late thought), Robert Donington17 has pointed out that there is no reason to think that any greater affinity existed between these two than between J. S. Bach and J. J. Quantz, who, as we have just observed, did not always agree with his colleague Emanuel. Quantz was only twelve years younger than J. S. Bach, was on friendly terms with him, and, with his links to musicians of the first rank who played the most up-to-date music in fash-ionable centres like Dresden, might have been a more sympathetic representative of the ‘new’ style. Certain ambiguous motifs in the Goldberg Variations permit contrasted interpretations, depending on how one
considers J. S. Bach’s own taste might relate to these two writers. The stylistic and cultural changes dur-ing and after Bach’s lifetime, and even the intangible nature of personal relationships, make it dangerous to grasp at evidence of Bach’s taste in the prescriptions of others—particularly those who reached their greatest fame after Bach’s death.18
The Goldberg Variations begin with a piece in the new galant style, of which both C. P. E. Bach and Quantz were champions. This style owed a great deal to French musical taste, but it would appear that Quantz was more wholeheartedly Francophile than Emanuel. For decades, the French had taken the single-note ornament very seriously. Their practice of detailed prescription would have appealed to J. S. Bach, who followed their custom of writing down an ornament table (illus.2). This, however, was merely a teaching aid for his young son Wilhelm Friedemann, and no such table appears in the context of any of Bach’s completed major keyboard works.19 The orna-ment table will receive further consideration below.
‘Optional’ ornaments?Bach annotated his own copy of the first edition of the Goldberg Variations (possibly with a second edition in mind), adding more ornaments than he had previously specified—but by no means in every place where they might have been appropriate.20 He expected more single-note ornaments to be added by the performer than he had indicated in the edi-tion, even including his annotations.
Quantz pointed out the necessity of adding appoggiaturas when they were not in the score.21 C. P. E. Bach went further, illustrating appropriate musical contexts for this.22 When annotating his own printed copy of the Goldberg Variations, his father added a series of appoggiaturas to Variation 26 in several of these contexts where he had ear-lier approved the print without feeling them to be necessary. This appears to contrast with Variation 23; however, Bach had to stipulate the succession of ornaments at the start of the second half, since they replace mordents used in the parallel situation dur-ing the first half (see ex.5, below).
The Versuch (Emanuel’s own examples fig.72h and fig.76) shows the reinforcement of a written-out appog-giatura by what C. P. E. Bach called the ‘short unvariable appoggiatura’ (Versuch, ‘Embellishments’, section 13: ‘It
Early Music PAGE 5 OF 16
5.5
5.10
5.15
5.20
5.25
5.30
5.35
5.40
5.45
5.48
5.50
5.55
5.60
5.65
5.70
5.75
5.80
5.85
5.90
5.955.96
carries one, two or three tails, and is played so rapidly that the following note loses scarcely any of its length’).23 J. S. Bach notated such reinforcements unambiguously in Variation 13 (ex.3, below). One can extrapolate from Bach’s explicit treatment here and apply this to paral-lel motifs within the same Variation—and elsewhere—and add more ornaments than the composer wrote.
Specified ornamentsBoth Emanuel and his father often wrote out longer appoggiaturas unambiguously in full note values. In the next example, however, they can be seen as com-monly notated. The four appoggiaturas in Variation 7 (ex.1a–b) are all indicated by small quavers. The
2 Bach, ‘Explanation of various signs, showing how to play certain ornaments correctly’ from Clavier-Büchlein vor Wilhelm Friedemann Bach
player can vary their length—and may indeed add many more.
In the next two examples (exx.2–3) the orna-ment must be played either slightly before the note to which it is applied, or virtually on the beat, but so short that it can equate with C. P. E. Bach’s pre-scription of the ‘short, unvariable appoggiatura’. The principal purpose must be to add weight to the note that follows, but Emanuel favours an ornament that just retains any dissonance that can be detected. (Importantly, if the ornaments in ex.3 are played as ‘real’ appoggiaturas, they would cause banal harmonies of a 6th and an octave respectively.) My preferred description for this type of ornament is a short, emphatic grace-note, whether played actually just on the main note, or slightly before it, the latter according more to French taste (see below) and probably to that of Quantz. Of course, in those instances where Quantz can be seen to be writing with the flute in mind, it is natural for him to think of an ornament played before the beat, since the flute cannot play two notes at once.
The type of ornament in exx.2–3 offers the harpsi-chordist a valuable means of emphasizing particular notes or motifs. The harpsichord defines the start of a note very clearly, but has limited dynamic control of individual notes. This makes it harder to control the effect of appoggiaturas, essential though this tech-nique is. The use of the short emphatic grace-note draws attention to a note played on this instrument, in a very useful manner—one less subtle than that achieved by delayed articulation. The reinforcement of an existing appoggiatura is particularly effective, although this, too, is a less subtle method than the slurred technique for performing appoggiaturas on the harpsichord, as described by Rameau.24
Variation 25 (ex.4) makes particularly expressive and varied use of single-note ornaments. Written-out mordents are embellished by an initial grace-note—akin to the practice of reinforcing a written appoggiatura. This may suggest a widespread, spontaneous use by Bach himself of grace-notes to give added interest to chosen notes or motifs.25 In Variation 25, if one tries to play these consistently as ‘short, unvariable appoggiaturas’ it is arguable
Ex.4 Bach, Variation 25, bars 1–3
Ex.2 Bach, Variation 24, bars 1–4
Ex.3 Bach, Variation 13, bars 1–4
Early Music PAGE 7 OF 16
7.5
7.10
7.15
7.20
7.25
7.30
7.35
7.40
7.45
7.48
7.50
7.55
7.60
7.65
7.70
7.75
7.80
7.85
7.90
7.957.96
that the lyrical quality of the melodic line is com-promised. The player must decide whether J. S. Bach would have thought a fractional forward placement of these ornaments to be unstylish or not: in other words, whether his taste here accorded more closely to that of his second son, or to that of Quantz.
Be this as it may, rising leaps throughout the piece generally carry appoggiaturas indicated by the same sign. As these resemble semiquavers and precede quavers, they may be interpreted as pro-portionate appoggiaturas—and usually are. But the aforementioned grace-notes which reinforce mordents are placed before demisemiquavers and are indicated by the same sign, which militates against such a literal interpretation. Again, Bach, who is generally consistent, writes out just two of these appoggiaturas as full semiquavers at a point of implied crescendo (during the second half, bars 21 and 22). Perhaps for most of the piece he expects either a different articulation, or the appoggiaturas to be generally shorter. A more varied treatment of them gives a more lyrical and spontaneous effect to the rhapsodic solo line.
The ornaments in Variation 23 (ex.5) are usually played as appoggiaturas—often proportionate ones. However, these will clash with the upward swoops in the left hand. Very short ornaments can avoid this, and enhance the playful mood of this startling piece. Here again, context must be the deciding factor.
The flexibility of the ornamentsI have questioned elsewhere whether Bach always achieved clarity in the area of rhythm.26 In my view, his obsession with a grammatically ‘clean’ score, and his reluctance to employ rhythmic ‘hints’ or other inconsistent devices, have left us with many ambiguous examples, where today’s performers can adopt very different treatments. But his ornamen-tation, too, remained far from definitive. Although
suffering contemporary criticism for being too elab-orate and specific in his notation of ornamentation,27 for much of his creative output he followed normal German practice for single-note ornaments, indicat-ing these sparingly—and often not at all. Nor did he remove some uncertainties regarding his intentions for some of those he did include, such as those in Variation 25 (ex.4) and Variation 23 (ex.5).
For those who seek concrete evidence of J. S. Bach’s intentions, the ornament table (illus.2) may seem to be a holy grail. The pedagogical reason for its existence has already been mentioned. However, it was not comprehensive, nor can one assume that Bach’s treatment of ornaments throughout his life was exactly as the table would imply. His ornament table bore a French title, and was a simplified adap-tation of d’Anglebert’s.28 In Bach’s, the single-note ornament retains d’Anglebert’s realization as a pro-portionate, half-length appoggiatura.29
An analysis of ornament tables30 will show that by Bach’s time the norm was to give only one realiza-tion of the single-note ornament—perhaps to avoid confusion. Many (but not all) composers presented the appoggiatura. Several reasons can be suggested. First, it is likely that the appoggiatura was gener-ally viewed as a more important ornament than an unstressed grace-note, and was the natural one to include. Second, it could more easily be realized in real notes, however simplistically. Finally, it was more modern: d’Anglebert may have been the first to feature it, in his edition of 1689. Furthermore, since C. P. E. Bach and Quantz were both to discuss situations where ‘proper’ appoggiaturas ought to be used even when not notated, we might conclude that for many composers, unstressed ornaments would even more commonly remain the prerogative of the performer.31
Nor should we infer from the French origin of Bach’s ornament table that he would have followed
Ex.5 Bach, Variation 23, bars 17–19
pAGE 8 OF 16 Early Music
8.5
8.10
8.15
8.20
8.25
8.30
8.35
8.40
8.45
8.48
8.50
8.55
8.60
8.65
8.70
8.75
8.80
8.85
8.90
8.958.96
7
13
17
22
27
31
3 Bach, Goldberg Variations, Aria: first edition (Nuremberg: Balthasar Schmid, c.1741), with bar numbers added
Early Music PAGE 9 OF 16
9.5
9.10
9.15
9.20
9.25
9.30
9.35
9.40
9.45
9.48
9.50
9.55
9.60
9.65
9.70
9.75
9.80
9.85
9.90
9.959.96
its directions for music of all styles. Ornament tables seldom define clearly their compiler’s inten-tions, and their interpretation can depend upon the user’s own origins and taste, and the musical con-text. Take, for example, an ornament within a falling 3rd—a frequent situation, but one absent from most ornament tables (ex.6). Often quoted in this context, the opening of the Flute Sonata in E major, bwv1035, is so modern, so galant, that it might have been writ-ten by a much younger composer. It may well be a self-conscious demonstration, written for the Berlin court, of Bach the father being perfectly able to write in the new style—when he chose.32 Quantz would have wanted the ornaments within the descending group in bar 1 to be played as grace-notes, before the beat; Bach’s son Emanuel would not. This ambigu-ity offers modern players two equally justifiable and very different modes of execution (neither of which would be based on mathematically proportionate appoggiaturas), each with subtle variants that defy attempts at precise notating.33 This brings us neatly back to the Aria of the Goldberg Variations, where the most striking ambiguities are to be found.
The AriaA comparison of illus.3 with ex.6 should encourage us to imagine the Aria played on the flute, and to assess the musical affekt of different treatments of its single-note ornaments in this light. Imitation of the subtleties readily produced by a good flautist would be a suitable (albeit difficult) objective of the good keyboard player.
However, an examination of the notation of single-note ornaments within the whole Aria also raises two important possibilities. Firstly, the sug-gestion that some of those ornaments indicated by small semiquavers might be played as nachschlage is of long standing. A century ago, Arnold Dolmetsch, who evaluated each ornament on the basis of
musical criteria alone, saw a nachschlag in bar 2 of the Aria (but only on the first beat). He interpreted the two identical signs in bar 2 as having quite dif-ferent functions.34 The result makes good musical sense, with a weak, linking ornament within the first beat, and a strong appoggiatura on the second. Dannreuther, in his even earlier work, had judged many ornaments in the Aria to be unstressed, and viewed both the ornaments in bar 2 as nachschlage.35 It may well be that these recommendations were in Ralph Kirkpatrick’s mind when he suggested a radi-cally different treatment.
Secondly, although quite recently examined by Beverley Scheibert,36 the idea seems inherent in Dannreuther’s work that Bach may have differen-tiated in his notation between two kinds of single-note ornament; this is worthy of some consideration. If the idea of differentiation is attractive, it would be a mistake to see Bach as entirely innovative in this respect—which in itself supports this concept, which may have been gaining ground for at least a decade. A similar use of two different signs was made by Rameau in his Nouvelles suites, probably composed shortly before their publication c.1729–30. The use in this volume of a small semiquaver to indicate an ornament shorter than that given as a small quaver was a new introduction to Rameau’s notational vocabulary. We also find this technique used by Georg Böhm in his D major ‘French-style’ suite. Böhm’s close, long-standing relationship to J. S. Bach is well known.
Even closer parallels occur in two fantasias by Telemann (from Fantaisies pour le clavecin; 3 dou-zaines (Hamburg, 1732–3)), another composer close to Bach, who would surely have owned a copy of them. In two slow movements from Set 2, the same two indications used by Bach occur in mark-edly similar musical contexts (exx.7–8).37 Short and longer appoggiaturas are clearly differentiated here.
Ex.6 Bach, Flute Sonata in E major, bars 1–2
pAGE 10 OF 16 Early Music
10.5
10.10
10.15
10.20
10.25
10.30
10.35
10.40
10.45
10.48
10.50
10.55
10.60
10.65
10.70
10.75
10.80
10.85
10.90
10.9510.96
Bars 8–9 of ex.7 offer a parallel to Bach’s bar 18 (see illus.3), where very short appoggiaturas lead to a longer one on the second beat. In ex.8 too we have a motif reminiscent of the Aria. The same ambigu-ity affects the placement of the short ornaments in the descending motif of bar 29. These two exam-ples are striking, and their galant modernity dates from some nine years before the publication of the Goldberg Variations.
The Aria: internal evidenceFor varying analyses of the potential execution of single-note ornaments on the basis of musical crite-ria, the reader is recommended to consult Dolmetsch and Dannreuther,38 where the idea of differentiation in Bach’s notation of these ornaments is considered to be worth exploring. This is appropriate, since it has either been given no consideration, or has been
consciously dismissed, by some present-day per-formers and scholars.39
Putting the matter simply, in the Aria most—one could argue, all—ornaments notated as semiquavers seem to indicate a lightweight ornament, and many may be played as some kind of grace-note, whereas it is hard to find any example of an ornament indi-cated as a quaver which might be played other than as an appoggiatura.40 Did Bach write here with the idea of differentiation only half-formed in his mind, and employ signs inconsistently? Otherwise, the player has to choose: either reject differentiation and apply the ornaments on the basis of musical criteria alone—or assume that Bach’s use of two different ornament-signs was conscious and purposeful, and that this may be some sort of guide to performance. In the latter case, some specific features of the whole work, and the Aria in particular, may be clarified.
Ex.8 Telemann, Fantasias, Set 2 no.4, bars 27–34
Ex.7 Telemann, Fantasias, Set 2 no.7, bars 1–12
Early Music PAGE 11 OF 16
11.5
11.10
11.15
11.20
11.25
11.30
11.35
11.40
11.45
11.48
11.50
11.55
11.60
11.65
11.70
11.75
11.80
11.85
11.90
11.9511.96
This would accord well with the prescriptive nature of Bach’s notation in general. Here is a short survey of those features:
(a) In bar 7 of the Aria, the right hand presents a reverse inequality of rhythm, notated in real note values. Since the second, third and fourth pairs constitute a downward motion of two falling 3rds linked by ornaments, Bach might be mak-ing a conscious contrast with a different place-ment of the ornaments in bars 2 and 6. (Those who prefer short appoggiaturas in bar 2 have to accept that Bach was using alternative notations to mean virtually the same thing, but delaying explicitness until the third instance, bar 7.)
(b) If the suggestion of contrasted rhythms in bars 2 and 7 is accepted, then the single-note ornament in bar 12 may also be placed before the beat, and the two ornaments in the first half-cadential bar, written differently, will be of different lengths.
(c) The ornaments in bar 18 of the Aria look the same as those already discussed, but in this case they begin on the first beat of the bar. They look like those in bar 2, but conventional ‘rules’ would dictate their performance as appoggiaturas. If differentiation by notation is in play, one might suggest that Bach wanted these to be very short and playful—weaker, indeed, than the ones writ-ten in real notes in bar 7, and offering another variation on the recurrent descending motif. One might also ask what other notational method Bach had, if it was actually his hope that in bar 18 these ornaments should be executed ‘playfully’, before the beat.
(d) In bar 18 the second beat carries the longer ornament-sign, unlike bars 2 and 6. This mili-tates against both Kirkpatrick’s and Dolmetsch’s interpretation of those bars. The stronger appoggiatura applied to the second beat of bar 18 must be intended to have more weight than the first beat. The very different feel which this simple change of sign might produce, between the openings of the first and second halves, would reflect Bach’s facility for increasing the force of the rhythmic (quasi sarabande) char-acter of the piece as it proceeds. Indeed, the weight which this treatment gives to that beat is
repeated by the application of similar appoggia-turas to the second beat of the next six bars (!), all indicated in the same way except for the fifth, where a lighter ornament has the effect of mak-ing the strong second beat of the next, the final bar in the sequence, even more climactic. The effect of this repetition of strong second beats could be that of a mild musical joke—possibly intentional, in the light of what can be consid-ered a descent into musical banality as the style of the composition changes completely in the last line.41
(e) Again depending upon differentiation, bar 26 may offer a comparison with the notational method of François Couperin, who had formal-ized an existing French ornament called accent, akin to a nachschlag.42 The purpose seems to have been to drag forward in the accompani-ment an ornament that might otherwise clash with one occurring in the more prominent part.43 Bach may be using two different signs to achieve the same effect here. The left-hand ornament can be played before the right-hand one, allowing the appoggiatura in the solo line to coincide with the bass d. A subtle difference is the slurring: Couperin slurred his accent to the preceding note, for greater clarity, whereas Bach continued to slur to the following note.
The Aria: conclusionsIf the single-note ornaments described above are seen as differentiated, Bach was following a rela-tively new, but natural notational trend—a move towards greater clarity and definition: one that was widespread, and which may have been in his mind, at least occasionally, for a number of years. In this case, we could then conclude that the ornaments indicated in the score by small quavers are all appog-giaturas, which may be played to varying lengths.
Those ornaments indicated by semiquavers offer a variety of treatments, but all of fundamentally similar lightness. This would make it unlikely that Bach wrote two in succession (as in bar 2), expect-ing a radically different weight to be applied to each. Some written in this manner (for example in bars 16, 18) may be regarded as appoggiaturas—and the dif-ferentiation in this case may indicate extreme light-ness within that definition, or might simply reflect
pAGE 12 OF 16 Early Music
12.5
12.10
12.15
12.20
12.25
12.30
12.35
12.40
12.45
12.48
12.50
12.55
12.60
12.65
12.70
12.75
12.80
12.85
12.90
12.9512.96
an observation of notational grammar. Others, whose function can be identified as emphasizing a following note, or as decorative links, may be played without emphasis. In some cases a placement before the beat may be appropriate. If one follows Quantz, the aim should be to make such ornaments sound as delicate and late as possible.44
Those who accept differentiation, but feel that C. P. E. Bach’s musical preferences must be closer to his father’s than those of Quantz, will favour var-ied appoggiaturas throughout the piece, with few or none of these occurring before the beat. In any case, though, it would have been widely considered in Bach’s day as unmusical for each half to be repeated in identical form: ornamentation would be liable to modification by the performer, although Bach’s own views on this are uncertain.
Beyond the implications of its time-signature and the note values within it, we cannot know whether Bach expected a particular tempo for the Aria, a choice that may radically affect the approach to
ornamentation. But if the rather simplistic conclu-sion of the piece suggests an intention to end in a light-hearted manner,45 this would be incompatible with a mood of solemnity at its start, or indeed with a performance offering the whole Aria in a spirit of sombre reverence—and with mathematically realized ornamentation. It is striking to compare the assess-ment of two scholars writing nearly half a century apart. Kirkpatrick wrote in 1938:The Aria seems to foreshadow the spirit of the whole work through the tenderness and calm with which the solemnity of the fundamental bass is clothed at its initial appearance.46
Dannreuther’s view, written at a time when per-formance of Bach was perhaps more objective, was this:The Aria is written in a style exceptional with Bach—that of an air à la mode, with a profusion of ‘galant’ ornaments, which must be treated after the manner of contemporary French chansons and German lieder, or like the embel-lishments in the slow movements of Quantz...47
The English harpsichordist and harpsichord-maker Colin Booth attended masterclasses given by Colin Tilney and others over several decades. Dartington International Summer School became an annual focus, and helped to inspire a dozen recordings of solo harpsichord music. At Dartington Colin led a series of seminars dominated by the subject of notational conventions, which was to result in a 2010 book for players, Did Bach really mean that? Deceptive notation in Baroque keyboard music. Discussions with leading performers on harpsichord, organ and piano had suggested a need for an approachable analysis of these conventions, which might reduce the tendency to literal perfor-mance among those trained in a modern tradition, and stimulate in keyboard players, whatever their chosen instrument, deeper understanding and greater freedom in performance. His own recording of the Goldberg Variations is on the Soundboard label. [email protected]
For invaluable help in the completion of this article, my thanks are due to Judith Lavin of the University of Toronto and Early Music’s anonymous readers.1 P. Badura-Skoda, ‘Let’s get rid of the wrong Pralltriller!’, Early Music, xli/1 (2013), pp.113–18.2 The ‘best modern editions’ may, of course, not present the full picture. Where the music is derived from a reliable autograph, or a good original edition (Telemann’s Fantasias; Graupner; Handel; Mattheson; Bach) we may draw conclusions. Many sources for music by Bach’s German contemporaries are derived from
copies—often made in Bach’s circle—which in some cases produced staff-based versions from tablature originals (for example, Reinken; Buxtehude; perhaps Böhm) where the indication of ornaments would be less likely. It may nevertheless be fair to draw conclusions regarding contemporary performance practice from these staff-based versions, even if details of the composer’s specific intentions may remain in doubt. The central question is whether a score that appears without significant decoration was meant to be played in this way; see K. Beckmann, Die Norddeutsche Orgelschule, Teil II (Mainz, 2009), ch.5.
As for J. S. Bach himself, an examination of his published works (compositions where we might expect to find greater care in the notation of such details) reveals a similar application to that of Graupner. Most single-note ornaments are to be found in slow pieces of an expressive nature. They are quite scarce in modern editions of the Partitas, but many are carefully stipulated in real note values, and some exceptions occur (for example, the Sarabande of the G major Partita no.5). In Clavierübung II, one element of the stylistic contrast between the Concerto in the Italian Style and the Overture in the French
Style is the difference in the frequency of single-note ornaments. They feature frequently in some movements of the latter, indicated mainly by small notes. The engraving of music was a laborious procedure, which led composers to indicate certain features of the music, like rhythmic detail and ornamentation, only at the start of movements. In general, Bach seems to have been reluctant to follow this practice, particularly in his published works. For those players who wish to get as close as possible to the original sources of the Goldberg Variations, the Performers’ Facsimile is a plain copy of the first edition, while the Fuzeau facsimile reproduces Bach’s Handexemplar (Bach’s own annotated copy), though in a form that is made less easy to use by a number of extra additions and a ‘smudgy’ reproduction.3 For example, ‘In the mid 20th century, few seem to have realised that Bach’s densely-written scores represent ornamented performance versions and therefore should be played in a flexible, quasi-improvised manner’; D. Fabian, Bach performance practice 1945–1975 (Farnham, 2003), p.139.4 The comments in n.2 concerning transmission through copies may also be applied to Bach’s own music, where the activities of the ‘Bach circle’ not only helped to preserve a lot of music by Bach himself, but inevitably introduced variants. Even a Handexemplar such as that of the Goldberg Variations, does not clarify Bach’s final intentions. Christoph Wolff has analysed this process with reference to much of Bach’s keyboard music, including corrections and additions that themselves present questions of authenticity. See C. Wolff, Bach: essays on his life and work (Cambridge, MA, 1991), chs.13, 15 and 16.5 Clavier-Büchlein Anna Magdalena Bach, 1725 (some entries undoubtedly later). Wolff presents the thesis that Anna Magdalena’s version of the Aria dates from around the same time as the first edition of the Goldberg Variations (1741), rather than pre-dating it, or even being by another composer; see Wolff, Bach: essays on his life and work, ch.13.
6 Reasons for this include: the long sustain of the modern piano, supposedly rendering trills unnecessary; the reduction over time of the amount of ornamentation, making it less appropriate to modern performance; and the view that Bach’s melodic lines actually sound better largely unadorned.7 R. Kirkpatrick (ed.), J. S. Bach—The Goldberg Variations (New York, 1938). Significantly, Kirkpatrick’s introduction discusses the instrument (the harpsichord is prescribed) as item (iii): after Origin (i) and Form (ii) and before Ornamentation (iv). As for the latter, although great weight is given to recommendations by C. P. E. Bach, the editor’s essay is balanced and informative, and for most players remains an admirable summary. In my view, it is the realization, which, by its selectivity and clear, continuous presentation on the page, has done most to lead to the rather unified performing tradition mentioned in this article.8 This is my personal judgement, based on a random but eclectic experience of performances, both live and recorded, mostly from the last few decades. A statistical assessment of live performance would be impossible. As for recordings, it is likely that performances became more, rather than less, standardized in the last few decades. For slightly earlier evidence, see Fabian, Bach performance practice, p.148, where it is pointed out that few ‘good’ editions existed in the 1930s, adding that ‘among harpsichordists, Kirkpatrick’s work would have been well known, and most importantly, in terms of executing ornaments there is relatively little divergence between the recordings, most playing as this score recommends’.9 E. Dannreuther, Musical ornamentation Part I (London, 1893–5), pp.202–3.10 A. Dolmetsch, The interpretation of the music of the xvii and xviii centuries (London, 1915), pp.153–4.11 W. Emery, Bach’s ornaments (London, 1953), pp.93–6.12 Kirkpatrick, J. S. Bach—The Goldberg Variations, p.xxvi: first
tempo, crochet = 62; second (18 months later) crochet = 56. But one comment (p.xxviii) is worth quoting: ‘The Aria seems to foreshadow the spirit of the whole work through the tenderness and calm with which the solemnity of the fundamental bass is clothed at its initial appearance’.13 Dorottya Fabian has analysed recordings of this music; see ‘Interpretation 1—tempo and dynamics’, in Bach performance practice, p.123.14 As Mendel put it 60 years ago, ‘Because of the essentially improvisatory character of trills, appoggiaturas, and other ornaments, the attempt to write out just what metric value each tone is to have can never be successful. I think this may be partly what Scheibe meant in criticising Bach for writing out so much … The attempt to pin down the rhythm of living music at all in the crudely simple arithmetical ratios of notated meter is (hardly) … possible.’ A. Mendel (ed.), Bach: St John Passion, vocal score (New York, 1951), p.xxii.15 Wolff, Bach: essays on his life and work, p.212, considers it more likely that Anna Magdalena Bach made her copy of the Aria from a manuscript, rather than from the first edition (see n.5). In this case, the differences in the values of some of the single-note ornaments within that copy may hypothetically reflect either, (a) the nature of the original from which she copied, implying that Bach refined the signs subsequently, or (b) her own opinion, that giving them specific values was unimportant. Yo Tomita offers a general, detailed study of Anna Magdalena as copyist in Understanding Bach, ii (2007), pp.59–76, online at www.bachnetwork.co.uk. The possibility that Bach made extensive use in a lost manuscript of his hook sign rather than small notes cannot be over-emphasized. Compare two versions of the Prelude in C! minor, Well-Tempered Clavier, Book II: the so-called ‘London’ autograph copy makes extensive use of hook symbols, whereas the copy from the ‘Bach circle’ by Altnickol uses small
notes. These can be examined in the Neue Bach Ausgabe, ed. A. Dürr, and a detailed study of the notation of this prelude has been made by Y. Tomita, J. S. Bach’s Das Wohltemperierte Clavier II: a critical commentary, ii (Leeds, 1993), pp.113–46. The same can be said of different manuscript sources of the French Suites.16 C. P. E. Bach, Essay on the True Art of Playing Keyboard Instruments [Versuch] (Berlin, 1752), ‘Embellishments: The Appoggiatura’, para.13; J. J. Quantz, On playing the Flute (Berlin, 1754), ch.xvii, part 2, para.6.17 R. Donington, Baroque music—style and performance, a handbook (London, 1982), p.3.18 Regarding the prescriptions of these figures (C. P. E. Bach, Quantz, Marpurg and others): ‘[none of] these solutions applies directly to the interpretation of J. S. Bach’s appoggiaturas. Philipp Emanuel … never pretends to be his father’s interpreter, nor does he even evince much interest in the masterpieces of the preceding generation’; P. Aldrich, Ornamentation in J. S. Bach’s organ works (New York, 1950), p.43. This may be considered an extreme view, considering, for example, the nature of C. P. E. Bach’s works published prior to his father’s death, and the publication date of both his and Quantz’s books, so shortly after that death in 1750. Both writers, for example, give simple, timeless instructions for the most basic kinds of appoggiatura, often to be performed proportionately, which it is hard to imagine J. S. Bach would have objected to. It is possible that Aldrich is in fact referring to ornaments that are not true appoggiaturas.19 Contained within the Clavier-Büchlein vor Wilhelm Friedemann Bach (1720).20 Variation 13, for example, could have an ornament added to the second beat of bar 24, and the last bar’s final right-hand c, echoing the motifs in bars 25–6.21 Quantz, On playing the Flute, ch.8, para.12.22 Bach, Versuch, ‘Embellishments: The Appoggiatura’, para.8–10, figs.70–2.
C. P. E. Bach goes on to illustrate the uses of the ‘short unvariable appoggiatura’ (figs.77–9), and finds more contexts for this than for the ‘variable’. Beginning with ornaments that appear in the score, he then seems to discuss suitable places for spontaneously introducing it (para.15).23 Bach, Versuch, ‘Embellishments’, para.13–15, fig.76a–g. He had already been drawn into giving an example of this ornament slightly earlier, at the end of his illustrations of the longer ‘variable’ appoggiatura (fig.72h). The difficulty of avoiding ambiguity is clear.24 Jean-Philippe Rameau, Pièces de Clavecin, Book ii (Paris, 1724), ornament table. Transcribed in C. Booth, Did Bach really mean that? Deceptive notation in Baroque keyboard music (Wells, 2010), p.33.25 Exactly as illustrated by C. P. E. Bach (see n.22).26 Booth, Did Bach really mean that?, ch.5, Part 2.27 For the criticism in print by J. A. Scheibe, and the subsequent defence by J. A. Birnbaum, see, among others, Donington, Baroque music—style and performance, p.95.28 Jean-Henri d’Anglebert, Pièces de Clavecin (Paris, 1689). Bach’s feelings in general about the French style are beyond the scope of this article, and are the subject of long-standing debate. As regards ornamentation, Hans Klotz has looked on Bach’s usage from a French perspective: H. Klotz, Die Ornamentik der Klavier und Orgelwerke von Johann Sebastian Bach (Kassel, 1984).29 This is the realization found in several ornament tables of this period. However, these are not great in number, and exceptions (Dandrieu, for example, who shows it as a before-the-beat ornament, together with Chaumont (1695) and Saint Lambert (1702)) should be noted.30 In addition to the survey contained in Booth, Did Bach really mean that?, ch.8, facsimiles of tables and analysis are presented by Klotz, Die Ornamentik der Klavier und Orgelwerke von Johann Sebastian Bach.
31 Slightly later, but probably reflecting a continuing practice, is the evidence of Nicolo Pasquali in his tutor of 1758: a realization is given to demonstrate common ornaments which shows the use of grace-notes to reinforce written appoggiaturas, even though none appears in the score. Nicolo Pasquali, The Art of Fingering the Harpsichord (London, 1758), examples reproduced in Booth, Did Bach really mean that?, p.307.32 It is likely that Bach composed the Sonata in E major in 1741 for Frederick the Great’s flute partner and chamberlain, Michael Gabriel Fredersdorf. Robert L. Marshall discusses the probable date of composition at some length in The music of Johann Sebastian Bach: the sources, the style, the significance (New York, 1989) pp.209, 220–4.33 The limitations of notation were stressed by Bach’s pupil Agricola in 1757: ‘It is not possible to establish exact rules as to what value the Vorschlag must have. It always remains somewhat arbitrary, depending on the feeling of the composer or performer.’ Noted by S. Babitz, ‘A problem of rhythm in Baroque music’, The Musical Quarterly, xxxvii (1952), p.559, and quoted in Fabian, Bach performance practice, p.144.34 Dolmetsch, The interpretation of the music of the xvii and xviii centuries, p.153.35 Dannreuther, Musical ornamentation Part I, pp.182–3.36 B. Scheibert, Jean-Henry D’Anglebert and the seventeenth-century clavecin school (Bloomington, IN, 1986), p.100.37 Telemann’s own published score of the Fantasias, on which the present examples are based, is online at http://imslp.org. The ornamentation is clearly indicated.38 Even the examples given earlier are sufficient to demonstrate the disparate nature of such interpretations, many of which have musical validity. It seems to me that the idea of differentiation of notation may go a long way to reducing these extremes of variety.
39 For example, David Schulenberg does not recognize differentiation. He has recently concluded that since bar 4 is a variant of bar 2, the appoggiatura in bar 4 implies an ornament of the same weight on the same beat of bar 2. I cannot see the validity of this argument. See D. Schulenberg, The keyboard music of J. S. Bach (London, 2006), p.378.40 To repeat: in my preferred terminology, ‘grace-note’ refers to an unstressed ornament. In performance this may be placed significantly before the principal note which follows it, or only slightly, or virtually on that note, being released very quickly. The appoggiatura is a stressed ornament, but it too may be extremely short, and thus executed with very little emphasis.
41 A view, we suspect, contrary to that of many. For example: ‘Starting with a quite high right-hand solo and moving towards more continuous and melodious semiquavers in the last half-dozen bars are characteristics that can be found in several of the sarabandes of the French Suites’. P. Williams, Bach—The Goldberg Variations, Cambridge Music Handbooks (Cambridge, 2001), p.54.42 François Couperin, Premier livre de pièces de clavecin (Paris, 1713). The accent (without, apparently, any need felt for a specific realization) is illustrated in the Explication des agréements, et des Signes. In her 1995 edition of L’Art de toucher le clavecin, Margery Halford reproduces the Table,
and gives two examples of the use of the accent.43 A protracted example from Les Baccanales (Ordre iv) is given in Booth, Did Bach really mean that?, pp.278–9.44 Quantz writes that ‘Short single-note ornaments [vorschläge], such as those occurring within intervals of thirds, must be executed very briefly and softly—as though, so to speak, only in passing’. On Playing the Flute, ch.xvii, part 2, para.20.45 See n.40.46 Kirkpatrick, J. S. Bach—The Goldberg Variations, p.xxviii (written in 1934).47 Dannreuther, Musical ornamentation Part I, p.202.
pAGE 16 OF 16 Early Music
16.5
16.10
16.15
16.20
16.25
16.30
16.35
16.40
16.45
16.48
16.50
16.55
16.60
16.65
16.70
16.75
16.80
16.85
16.90
16.9516.96
Abstract
Colin Booth
Bach’s use of the single-note ornament in the Goldberg VariationsThe single-note ornament was indicated by German composers of Bach’s time neither as often, nor so specifically, as by French composers of the period. Among German authorities, both C. P. E. Bach and J. J. Quantz emphasized the fact that single-note ornaments were to be found expressly indicated far less often than a good performance would demand. C. P. E. Bach detailed numerous contexts where they ought to be applied even when not written.J. S. Bach’s Goldberg Variations present an example of a composer being more prescriptive than was normal in his use of this type of ornament. Some of the Variations are consciously and more-or-less fully ornamented, a process which was pursued fur-ther by Bach in his annotation of his own copy of
the first edition. Study of the work as a whole, and of these movements in particular, can assist players to achieve a suitably stylish performance not only of those Variations where Bach’s indications remained incomplete, but of pieces in comparable styles where ornaments were indicated only sparsely or not at all.The single-note ornament itself is liable to a vari-ety of interpretations, even when indicated in the score. This article surveys the principal musical uses of the ornament, with reference to the Goldberg Variations. It concludes by discussing the sugges-tion that Bach differentiated between two types of single-note ornament in his notation of this work—a practice that may have been gaining ground among leading composers from around 1730, in more than one country.
Keywords: J. S. Bach; Goldberg Variations; ornaments; C. P. E. Bach; J. J. Quantz; appoggiatura; grace-note; Nachschlag
![Variations-Goldberg [BWV 988] - Sheet music · Title: Variations-Goldberg [BWV 988] Author: Bach, Johann Sebastian - Arranger: Busoni, Ferruccio - Publisher: Leipzig: Breitkopf &](https://static.documents.pub/doc/80x56/613b27b8f8f21c0c8268d813/variations-goldberg-bwv-988-sheet-music-title-variations-goldberg-bwv-988.jpg)
